On August 3, local time, the trade secret dispute between Apple and OpenAI took a significant turn. That day, Apple formally filed a motion for a preliminary injunction with the U.S. District Court for the Northern District of California, requesting that OpenAI and two former employees involved in the case be barred from accessing, obtaining, using, or disclosing the confidential information Apple claims is at stake, while also seeking to expedite the discovery process.
Just hours later, OpenAI fired back with a public response on its official website titled “Apple is getting this wrong”, flipping the script by releasing early email correspondence between the two companies and iMessage chats involving the accused employees. In it, OpenAI flatly denies the core allegations, insisting the lawsuit is built on misinformation and incomplete facts.
The friction between these two tech giants dates back to July 10, when Apple initially filed suit in the same Northern California court. The defendants include OpenAI, its hardware subsidiary io Products, and two former Apple engineers — Chang Liu and Tang Yew Tan. Apple alleges that these two individuals carried unreleased product details, R&D processes, and supply chain trade secrets into OpenAI for use in its consumer hardware development projects. Apple further claims that OpenAI facilitated the transfer of confidential information by poaching employees and turning a blind eye to improper handling of former employer data.
Apple’s complaint spells out several specific accusations. For one, Liu allegedly failed to return all work devices upon departure and exploited a loophole in Apple’s internal system permissions, maintaining access to the company’s cloud storage for weeks after leaving. During that window, he reportedly downloaded dozens of high-level engineering files — including thousands of pages of manufacturing test protocols for main logic boards. For another, Apple asserts that OpenAI has systematically targeted its talent pool, with over 400 ex-Apple employees now at OpenAI as of the filing, spanning core roles in hardware design and supply chain management. Apple argues this isn’t normal market hiring churn but a deliberate strategy aimed at extracting proprietary know-how.
Following the initial lawsuit, OpenAI issued brief statements on July 10 and July 14, saying it has zero interest in other companies’ trade secrets and has seen no evidence supporting Apple’s claims.
In its latest response, OpenAI takes aim at Apple’s narrative about pre-litigation communication. Apple had claimed in its lawsuit that it sent a letter to OpenAI as early as February 2026 about these concerns and received no reply, painting OpenAI as dismissive of compliance issues. But OpenAI’s released email records tell a different story. According to OpenAI, Apple’s outside counsel mixed up two Asian surnames and mistakenly sent the key correspondence to an unrelated person. The so-called call with OpenAI’s general counsel never actually happened. When OpenAI pointed out the mix-up, Apple’s attorneys acknowledged the name confusion but never mentioned any specific allegations from the future lawsuit — instead, they reportedly told OpenAI the two sides were “resolving all issues.”
For the next five months, OpenAI says it heard nothing further from Apple until the lawsuit landed. OpenAI contends that Apple deliberately buried its own communication blunder and painted a seriously distorted picture of the early discussions in its legal filings.
As for the trade secret theft accusations, OpenAI released redacted iMessage transcripts. According to those chats, Liu’s file sharing on his departure date of January 22, 2026 — and in the days after — wasn’t about feeding secrets to OpenAI. Instead, current Apple employees were actively asking him to help locate files, provide project details, and share contact info for colleagues, all to facilitate internal Apple work handovers.
The transcripts show Apple staff using a device still logged into Liu’s personal iCloud account to copy files before finally logging out. Even weeks after his departure, former colleagues were still reaching out to him for project specifics and engineer contacts. Based on this, OpenAI argues that what Apple calls “confidential information leaks” largely stems from its own sloppy offboarding procedures and broken internal transition workflows — not from employees actively handing over trade secrets. OpenAI also emphasizes that Tan, upon joining, explicitly instructed his team to avoid obtaining or using any confidential information from prior employers. The company states it “neither possesses nor wants any Apple trade secrets.”
Beneath this legal tug-of-war lies a deeper strategic collision: OpenAI is accelerating its push into consumer AI hardware, directly encroaching on Apple’s core turf. In 2025, OpenAI acquired hardware firm io Products for $6.5 billion, officially stepping into the smart device arena. Yet just a year earlier, market speculation was rife that Apple and OpenAI would partner to integrate ChatGPT into iOS, shoring up Apple’s AI capabilities. But as both companies’ strategies have evolved, their business territories now overlap fast. Apple is doubling down on its Apple Intelligence initiative, deepening the fusion of in-house large models and end-device hardware. OpenAI, meanwhile, is extending from pure software algorithms into physical hardware, chasing what it hopes will be the next gateway in the AI era.
